With the goal of finding serological markers to monitor patients with early- as well as late-stage melanoma, we compared the levels of the cytoplasmic melanoma-associated antigens (CYT-MAA) and high-molecular-weight melanoma-associated antigen (HMW-MAA) in the sera of melanoma patients and controls. Using double-sandwich ELISA, we measured levels of both antigens in 117 patients and in 62 age- and sex-matched controls. Patients were stratified into four risk group based on stage of the disease. Serum levels of both markers were significantly higher in melanoma patients than in controls. CYT-MAA was the more sensitive marker, with 61% of patients showing elevated levels regardless of the stage of disease. HMW-MAA was elevated in 29%. Elevated CYT-MAA was also significantly correlated with poorer clinical outcome. By multivariate analysis (adjusting for stage and age), patients who had elevated CYT-MAA were 81% more likely to recur than patients with undetectable levels (hazard ratio=1.81, 95% CI=[1.07, 3.06], p-value=0.03). Elevated levels of HMW-MAA did not correlate with poor prognosis. These results suggest that both CYT-MAA and HMW-MAA are serum markers for residual melanoma in patients with resected disease. Furthermore, CYT-MAA appears to be a prognostic marker of clinical outcome in melanoma vaccine-treated patients.

Reduced DNA repair has been linked to an increased risk of cutaneous malignant melanoma, but insights into the molecular mechanisms of that link are scarce. The INK4a/ARF (CDKN2a) locus, which codes for the p16(INK4a) and p19ARF proteins, is often mutated in sporadic and familial malignant melanoma, but it has not been directly associated with reduced DNA repair. We transfected unirradiated mouse fibroblast cells with UV-treated DNA to measure DNA repair in normal, p16INK4a mutant, p19ARF mutant, or double mutant mouse host cells. Loss of either p16(INK4a) or p19ARF reduced the ability of the cells to process UV-induced DNA damage, independent of cell cycle effects incurred by the loss. These results may further explain why INK4a/ARF mutations predispose to malignant melanoma, a UV-induced tumor.

Simple, noninvasive methods are needed to follow effectiveness of new treatments in patients with melanoma. In our study, we examined cytoplasmic melanoma‐associated antigen (CYT‐MAA) serum level in melanoma patients during immunotherapy. Sera of 117 patients were assayed for CYT‐MAA by double‐sandwich ELISA before and during treatment with a polyvalent, shed antigen, melanoma vaccine. Vaccine‐treated patients included 30 with American Joint Committee on Cancer (AJCC) stage IIb or IIIa, 30 with stage IIc, IIIb or IIIc, 30 with resected stage IV and 27 with measurable stage IV disease. Prior to vaccine therapy, 63% of patients had elevated serum CYT‐MAA with high levels of antigen in all disease stages. After initiation of therapy, the level declined in more than 90% of the positive patients and fell below the positive cut‐off in 56% of these patients within 5 months. By contrast, there was no decline in CYT‐MAA serum level in 11 patients who served as untreated controls with melanoma. Multivariate analysis of the treated patients using accelerated failure time Weibull models adjusted for stage and age showed that patients whose CYT‐MAA serum level remained elevated during treatment were ∼3 times more likely to recur or progress than patients who were consistently below the positive cut‐off (hazard ratio = 3.42, 95% CI [1.38, 8.47], p = 0.0079). Measurement of CYT‐MAA serum level appears to show potential as an early marker of prognosis in patients with stages IIb to IV melanoma. Measurement of CYT‐MAA serum level during therapy could provide an intermediate marker of response in these patients. © 2006 Wiley‐Liss, Inc.

No abstract

The Ink4a/Arf (CDKN2a) locus encodes two proteins that regulate two of the most important tumor suppressor pathways represented by p53 and Rb.(1) Loss of either p16(INK4a) or p19(ARF) was recently reported to reduce the ability of mouse cells to repair UV-induced DNA damage and to induce a UV-mutator phenotype. This observation was independent of cell cycle effects incurred by either p16(INK4a) and/or p19(ARF) loss, as it was demonstrable in unirradiated cells using UV-treated DNA. We suggest that this might explain why germ line mutations of INK4a/ARF predispose mainly to malignant melanoma, a UV-induced skin cancer, and provides a molecular explanation for the link between melanomagenesis and impaired DNA repair. It also further demonstrates that regulation of cell cycle check points and DNA repair in response to genomic insults, such as ultraviolet irradiation are intricately interwoven processes. Differences in the apoptotic response to ultraviolet light between melanocytes and keratinocytes might explain why INK4a/ARF mutations predispose to malignant melanoma, but not to keratinocyte-derived skin cancers.